Determining the location of objects via a camera system has many advantages. Prior attempts to locate objects have required the use of two spaced-apart cameras trained at the object. Use of two views allows for depth perception.
Additionally, determining the location of particular objects can provide the foundation for additional applications. For example, a determination of the location of a particular vehicle in a large parking lot can be quite beneficial to the driver that forgets where they parked.
It is therefore a challenge to develop a method to track images to overcome these, and other, disadvantages.